Improvements in Real-Time Feature-Based License Plate Character Recognition

Abstract

This paper presents an effective method for segmenting and recognizing license plate characters on a moving vehicle in actual outdoor environment. It is applicable to various camera conditions and installation environment and provides some effective methods to improve the rate of recognition and procedure speed. This paper is concerned not only implementations in laboratory environment but also in outdoor environment like roads, entrances of parking areas and other similar places where license recognition is needed. The digital images used can be obtained from any digital camera. The proposed vehicle license plate character recognition methods are constructed from a feature-based approach. By regulating a character, a unique structural feature for each character can be extracted. The extracted feature is insensible to changes in the outdoor environments. The proposed recognition system has two main processes: 1) to construct the feature information and 2) to segment a character image from any object image of a license plate and identify it. The relative regulations, the similarity and dissimilarity between each character, are used as recognition features.